Subject: Shift report 28-Aug/01-Sep-2000
Nicola.Zaccheo@lngs.infn.it
Date: Fri Sep 01 2000 - 11:57:46 EDT
Data shift report (28-Aug-2000 to 01-Sep-2000)
--------------------------------------------------------------------------
by Nicola Zaccheo
From run 19867 to 19890
--------------------------------------------------------------------------
..Not a good week for MACRO...
First problems during last week-end and last Monday:
the streamer tubes were firing a lot on SM4, SM5 and SM6 expecially
during run 19870 (26-Aug)...In this run I found a high TOHM trigger
rate (~ 10mHz) on all the SM5 planes and a high LIP trigger rate for some
boxes on SM5....I think this was induced by the noise generated by
the streamer planes (?)..(The situation retrurned normal after some
streamer planes have been turned off).
During run 19867 (the run restarted by Yiannis from the USA) the tank
5C02 had a crazy TOHM trigger rate (~1 Hz)..After this run it returned O.K.
..I think 5C02-0 is now working with one PMT only (remember my last shift
report?)...in fact, from run 19869, we are receiving some PHRASE L/R
warnings regarding this counter that has now L/R=0.53..Unfortunately I had
no time to run the calibration code for this week and so I can not
make a cross check of this using ERP.
The 5C02-0 signal (I think from one PMT only) looks now better than
in the past week. In addition there are no longer "fake" latches
on the STPM CIT SM5@CENT..(my last week theory was correct..)
We received two DSP warnings (27-Aug and today) regarding M:2,B:2,C:4..
..I found the +12V giving +10.9V only (w/o any AC components)..I tryied
to increase this voltage value using the voltage adjustment on the
regulator card w/o success...We received the second DSP warnings while
I was working on this...It reported problems for -24V (found -21.6V) too,
but this was not real (I found exactly -24V on -24V line).
Note that during the run in wich this power supply had lower voltage
on +12V, the LIP trigger rate on SM4 was lower than normal.
All the MACROUSA run analysis jobs paused because of lack of disk space on
DISK$SCRAUSA6. In addition, I found the PREP_WWW job still running since
last Friday (at that time there was a AXPGS5 boot). I stopped it with
Nazzareno's help and I restarted it.
On Tuesday:
I found the 1500 Kinetic power supply on M2,B2,C4
(the one that gave the EMILY DSP warnings) with no +12V/+24V.
I started a new run w/o uVax2.
I replaced it with the same supply that was on the (ex)calibration crate,
but this last one lost the -12V with full load (i.e. with both TOP *AND*
WEST/EAST attico TOHM modules) but it worked with a partial load
(i.e. TOP *OR* W/E TOHM attico).
I tried also turning off the PMTs fanout but this p.s. still didn't start
with both TOHM modules.
I tested the power supply that we replaced (from the same crate) after last
MACRO shut down and it lost -12V/-24V with a 4A load.
On Wednesday:
I found run19880 crashed (at ~6.10AM) for a broken PHRASE fan-unit on SM2.
I replaced it with a spare one and I restarted a new run.
Then I fixed the 1500 Kinetic power supply that yesterday I removed from the
TOHM SM4 crate. I replaced the positive voltage regulator card.
After ~ 1hr of mini-acq (that I analysed) I put back in acq uVax2.
The first run with the full detector is run 19882.
On Thursday:
this night run 19883 stalled. Yiannis restarted a PHRASE run only
from the USA. This morning I again found the Kinetic 1500 power
supply on B:2,C:4 (SM4) w/o all the positive voltages. Checking
the positive voltage regulator card, I found a 10Ohm resistor (R9
in the schematic) fused(!). I replaced it, I tested the power supply
(with 3A and 4A load) and it was O.K., then I put back it on the crate.
When I turned it on, after loading the CAMAC list, the West/East TOHM
attico module had all the red and green LEDs (act-0, act-1 and trigger)
of channels 7 and 11 stuck, even if I unplugged the input signals (note
that no signals were present from the trigger output and that the power
supply continued to work fine). Furthermore, it was impossible to select
channels from the key on the front-panel of the module. I removed this
module, I checked it w/o finding anything of trivial, then I put it back
on the crate...no red lights lighted (the channel select still didn't work).
I performed a mini-acq for more than 1hr (1000 events) and, analysing
the data, all the SM4 TOHM attico west channels were died while all the
SM4 TOHM attico east ones had the same high trigger rate (~80mHz)..
At this point I touched the TOHM module and then *all* the LEDs lighted...
I don't know if the power supply breaking caused the module breaking or
'viceversa'...there is something of strange on the electrical power on
the SM4 (problems for the streamer tubes there, fire-alarm Saturday night,
the uVax2 VT420 terminal found broken yesterday...)
at this point I removed the SM4 TOHM Weast/East attico module from the crate
and I prepared a CAMAC list in which I excluded the TOHM West/East
attico. I didn't remember how to pad camac zeros in the data stream..
(I was looking for Sophia to ask it to her but she had some problems
and she was at LNGS in the late afternoon), so I left the module readout
uncommented in the camac list otherwise there was a swapping
between W/E attico <-> TOP face (equip 27 read W/E attico before TOP) and
I started a new run with uVax2 back in acquisition.
Today:
I excluded uVax2 in order to test the camac list with the
"WRITEHEX $000000" (Sophia dixit) at the place of the redaout of the
module removed yesterday and everything went fine
(the 21,22,23 and 24 words of the 28 uVax2 words were always zero).
So I put back uVax2 in acquisition starting from run 19892 with
this new camac list.
During run 19889 there were some "machine check" alarms from
MV6. I stopped the run and replaced a PHRASE power supply on
crate B of SM1 Rack 9. Then I started a new run.
The MACRO Pirate Web Server batch job (MPWS) was not running
this morning (the MACRO On-line web page was not available).
I contacted Nazzareno and he told me that WSGS02 has now a new VMS
(VMS 7.2) and Multinet (4.1) version installed. He suggested that all
the Cuck's executable should be re-linked under the new VMS 7.2/Multinet 4.1.
(even if it's maybe sufficient only to restart the LANE'S jobs).
Checking the data, I found a very low rate from the ERP Bari box
on SM5. I checked the AMP connector thet seems to be fine. I had
to investigate this better the next week.
TRIGGER RATES:
TOHM --> 4T13 ~8mHz, 5B13 ~6.5mHz
ERP --> 3B07 ~6.5mHz
LIP --> 1B11 ~1.3mHz, 1B12 ~0.7mHz, 1B14 ~0.8mHz, 1B16 ~0.2mHz(!)
1C08 ~1.4mHz, 1C15 ~1.2mHz, 1T09 ~1.4mHz, 1T10 ~1.4mHz,
2E10 ~17.5mHz, 3E04 ~7.2mHz, 5T02 ~0.6mHz, 5T03 ~0.8mHz,
6C09 ~1mHz, 6C10 ~1mHz, 6T12 ~11mHz
ERP-GC --> 6W12 ~95mHz, 6S01 ~60mHz, 6S04 ~92mHz, 6S07 ~90mHz.
PHRASE --> 5B04 ~30mHz, 6E12 ~25mHz (3T12 ~55mHz during run 19881)
FMT:
No problems reported
LAMOSSKA:
There are still not justified LAMOSSKA triggers for LAM50.
CSPAM/ERP efficiency:
The boxes with systematic inefficiency: 3B07 (~48%), 3B12 (~70%),
5W03 (~56%) and 6T07(~70%).
ERP/LIP efficiency:
Low efficiency for 1B08, 1B09, 1B10, 1B12, 1B16, 1C08, 1C11, 1C15, 1T10,
1T13, 1T16, 4C08, 5C11, 5C13, 5T02, 5T03, 6B07, 6B10, 6C09, 6C10 and 6T11.
SPU: still hits on ERPI, ERPII, SPARE, LAS and LED on uVax1 (sometime 1/2
VETO bits too).
STOP MASTER:
"Fake" latches on SM5@CENT no longer present (see previous comment).
This archive was generated by hypermail 2a24 : Fri Sep 01 2000 - 11:57:49 EDT